About

Shufa Du, MD, PhD, is a nutritional epidemiologist at the UNC Gillings School of Global Public Health, serving as a professor in the Department of Nutrition. His research focuses on how underlying factors such as education, income, and genetics influence dietary behaviors and physical activity or inactivity, and how these behaviors impact health outcomes including cardiovascular diseases, diabetes, and obesity. Dr. Du has been involved in an international collaborative cohort study, the China Health and Nutrition Survey, and has served as a mentor on an NIH Fogarty training project related to nutrition-related non-communicable diseases prevention in China since 1996. He maintains a long-term collaborative relationship with the China Centers for Disease Control and Prevention, which supports students' global fieldwork internships.

Research topics

• Medicine
• Environmental health
• Evolutionary biology
• Biology
• Genetics
• Gerontology
• Endocrinology
• Demography
• Computer Science
• Geography
• Internal medicine
• Psychiatry
• Psychology
• Virology
• Pediatrics
• Computational biology
• Bioinformatics

Selected publications

• Snack preferences among school-aged children: evidence from discrete choice experiments

  BMC Public Health · 2026-03-09

  articleOpen accessCorresponding

  Accelerated economic development and innovative food industry trends have contributed to worsening dietary imbalanced among Chinese school-aged population. To investigate snack preferences among school-aged children, determine the relative importance of different snack attributes, and explore how these preferences vary by social-demographic characteristics to inform targeted nutritional interventions. A stratified cluster random sampling strategy was employed to recruit 854 school-aged children (grades 7 and 8) from Hubei and Anhui provinces. An evaluation framework comprising six attributes (taste, nutrient claims, purchase location, price, package size, and social influence) was constructed based on a Discrete Choice Experiment (DCE). A D-optimal design was implemented using SAS 9.4 software to generate 16 choice sets (divided into two versions), with two-stage questions incorporated to mitigate bias. A mixed logit model was applied to calculate preference coefficients (β), relative importance (RI), and willingness-to-pay (WTP) for different attributes. All six included attributes demonstrated statistically significant effects on local school-aged children’s snack choices (all P < 0.05). Specifically, school-aged children showed stronger preferences for snacks sold in supermarkets (β = 0.439), those rich in dietary fiber (β = 0.611), with sweet taste (β = 0.471), and commonly consumed by family members (β = 0.452). They were willing to pay an additional 46.49 CNY for snacks rich in dietary fiber. Large package size (β = -0.112) and price (β = -0.013) showed slight negative associations. Heterogeneity analysis revealed that boys preferred snacks rich in dietary fiber (β = 0.554) and were more price-sensitive (β = -0.026), while girls prioritized low-fat options (β = 0.738) and showed lower willingness to choose large packages (β = -0.259). School-aged children with less-educated parents placed greater emphasis on taste (RI = 25.33%), whereas those from highly-educated families valued package size more (RI = 10.43%). Overweight/obesity school-aged children preferred spicy snacks (β = 0.448) and were more price-sensitive (β = -0.115), while normal-weight school-aged children tended to choose sweet-tasting snacks (β = 0.510). Snack preferences among school-aged children vary significantly by socio-demographic characteristics. Interventions should be tailored to the specific characteristics of them, while simultaneously engaging both the school-aged children and their parents to enhance participation and educational effectiveness, thereby reducing the purchase and consumption of unhealthy snacks.

  PublisherDOI

• From Words to World: Attitudinal Appraisal Analysis of China's Anhui Provincial Tourism Branding on TikTok

  LEKSEMA Jurnal Bahasa dan Sastra · 2026-01-14

  articleOpen access1st authorCorresponding

  Digital media has opened new avenues for emerging destinations to shape tourist perceptions by projecting their tourism imaginaries to global audiences. However, there is a lack of study on how tourism imaginaries are constructed through attitudinal resources, especially in the context of emerging destinations. Therefore, this study examines the discursive construction of tourism imaginaries in Anhui Province, China, through its official TikTok promotions. A corpus of 203 videos was compiled, and the verbal elements were analyzed using corpus-assisted discourse and semantic analysis. Anhui’s tourism imaginaries and the attitudinal resources based on the attitude subsystem of Appraisal Theory were identified. The evaluative meanings embedded in these imaginaries were then examined through qualitative discourse analysis. The findings show that Anhui mainly uses ‘appreciation’ and ‘affect’ resources to construct positive tourism imaginaries, which include majestic natural scenery, timeless elegance, vibrant cultural festivals, exquisite cuisine, and dynamic urban life. This study suggests that the attitude system offers a useful lens for examining the verbal construction of tourism imaginaries in digital promotional contexts and demonstrates how evaluative language can be strategically used to elevate destination branding and enhance global visibility.

  PublisherOA PDFDOI

• Long-term body mass index trajectories and dyslipidemia and related gut microbial features: results from the China health and nutrition survey 1991–2018

  BMC Public Health · 2025-09-30 · 1 citations

  articleOpen access

  OBJECTIVE: To discover the associations between long-term body mass index (BMI) trajectories and dyslipidemia, and the internal mechanism within different BMI trajectory populations. METHODS: We chose 11,499 adults from China Health and Nutrition Survey (CHNS) for trajectory modeling. Selected those with two blood lipid measurements (n = 2,003), and determine the risk trajectory group associated with dyslipidemia. We selected two subsets of the 2015 CHNS, CHNS-1 (n = 3,061) and CHNS-2 (n = 1,409) for gut microbiome analysis. A Lasso regression in both the CHNS-1 and the CHNS-2 confirmed the dominant genera of each dyslipidemia risk trajectory. We also used a subset of CHNS-2 (n = 779) with metabolomics data to identify the differential metabolites. We ascertained the connection between microbiome and metabolites via correlation analysis. RESULTS: = 0.50 and 0.23, p < 0.001. we found six taxa in the OTO group and five taxa in the DO group had lower abundance than in the NS group. after adjusting for covariates, yielded an AUC of 0.759 and 0.646 in the validation set. In the OTO and DO groups we also found 36 metabolites and 5 metabolites, respectively, with differential concentrations than in the NS group. CONCLUSIONS: A high-level BMI change trajectory will increase the risk of developing dyslipidemia in the future. The growth trajectory of BMI affects lipid metabolism mediated by gut microbiota.These findings may reveal the mechanism of BMI changes leading to dyslipidemia, and indicate the corresponding microbiome targets for prevention.

  PublisherOA PDFDOI

• Towards the Data Analysis of Fibonacci Sequence During SSE 50ETF Options Trading in China

  Lecture notes in computer science · 2025-01-01 · 1 citations

  book-chapter
  PublisherDOI

• Association of allostatic load and dietary inflammatory index with depressive symptoms among U.S. adults: NHANES 2007–2018

  Journal of Affective Disorders · 2025-07-18 · 3 citations

  article
  PublisherDOI

• Spatiotemporal Evolution of Carbon Storage and Driving Factors in Major Sugarcane-Producing Regions of Guangxi, China

  Agronomy · 2025-07-27

  articleOpen access

  Objectives: The major sugarcane-producing regions of Guangxi represent a critical agricultural zone in China. Investigating the mechanisms of land use change and carbon storage dynamics in this area is essential for optimizing regional ecological security and promoting sustainable development. Methods: Employing the land use transfer matrix, the InVEST model and the Geodetector model to analyze carbon storage changes and identify key driving factors and their interactive effects. Results: (1) From 2011 to 2022, Guangxi’s major sugarcane-producing regions experienced significant land use changes: reductions in cultivated land, grassland and water bodies alongside expansions of forest, bare land and construction land. (2) The total carbon storage in Guangxi’s major sugarcane-producing regions has increased from 2011 to 2018 by 0.99%, representing 1627.03 and 1643.10 million tons, while it has decreased by 0.1% in 2022 (1641.47 million tons) compared to 2018. (3) Cultivated land proportion and forest coverage rate were the primary drivers of spatial heterogeneity, followed by average slope and land urbanization rate. (4) Interaction analysis revealed strong synergistic effects among cultivated land proportion, forest coverage rate, NDVI and average slope, confirming multi-factor control over carbon storage changes. Conclusions: Carbon storage in the Guangxi sugarcane-producing regions is shaped by land use patterns and multi-factor interactions. Future strategies should optimize land use structures and balance urbanization with ecological protection to enhance regional carbon sequestration.

  PublisherOA PDFDOI

• A randomized controlled trial of acupotomy lysis with Tiaoji Tongdu for the treatment of cervical spondylotic radiculopathy

  Lishizhen medicine and materia medica research · 2025-07-01

  article

  <sec> <title>目的</title> 探讨调脊通督针刀松解术治疗神经根型颈椎病（CSR）的临床效果及安全性。 </sec> <sec> <title>方法</title> 将符合纳排标准的108例CSR患者随机分为试验组54例和对照组54例，因对照组脱落1例故最终纳入107例。试验组予以调脊通督针刀松解术治疗，对照组予以常规针刺治疗。记录并比较两组患者在治疗前、后VAS评分、NDI指数、YT20SS评分，测定炎症因子TNF-α、IL-1β、IL-6水平，观察不良反应，并在治疗结束第12周时随访。 </sec> <sec> <title>结果</title> 两组患者VAS评分、NDI指数、炎症因子TNF-α、IL-1β、IL-6水平均较前降低（<italic>P</italic>&lt;0.05），且试验组上述指标均低于对照组（<italic>P</italic>&lt;0.05）；两组患者YT20SS评分较前升高（<italic>P</italic>&lt;0.05），且试验组高于对照组（<italic>P</italic>&lt;0.05）；试验组总有效率（96.30%）高于对照组（83.02%），复发率（9.26％）低于对照组（28.30％）。差异有统计学意义（<italic>P</italic>&lt;0.05）。两组患者均未见明显不良反应。 </sec> <sec> <title>结论</title> 调脊通督针刀松解术和常规针刺均能改善CSR患者的疼痛麻木症状及功能障碍，并降低炎症反应，但调脊通督针刀松解术的治疗效果更显著、持久，值得临床推广使用。 </sec>

  PublisherDOI

• Spatiotemporal Evolution of Carbon Storage and Driving Factors in Major Sugarcane-Producing Regions of Guangxi, China

  Preprints.org · 2025-06-13 · 2 citations

  preprintOpen access

  [Objectives] The major sugarcane-producing region of Guangxi represents a critical agricultural zone in China. Investigating the mechanisms of land use change and carbon storage dynamics in this area is essential for optimizing regional ecological security and promoting sustainable development.[Methods] Based on four-phase land use data (2011–2022), this study employed a land use transfer matrix and the Geodetector model to analyze spatiotemporal changes in carbon storage (assessed using the InVEST model) and identify key driving factors and their interactive effects.[Results] (1) From 2011 to 2022, total carbon storage in the study area fluctuated between 1,627.03 and 1,644.17 million tons, exhibiting a northwest-high, southeast-low spatial pattern, with high-value zones concentrated in mountainous regions and low-value areas in economically active lowlands. (2) Land use patterns significantly influenced carbon storage: forests remained the dominant contributor (&amp;gt;85% of total storage), while cropland and bare land initially declined before recovering. Grassland and water bodies showed sustained carbon loss, whereas construction land expansion drove carbon increases. (3) Land urbanization rate (q &amp;gt; 0.647) and cropland proportion (q &amp;gt; 0.656) were the primary drivers of spatial heterogeneity, followed by nighttime light index and forest coverage (q &amp;gt; 0.511). (4) Interaction analysis revealed strong synergistic effects among NDVI, forest coverage, and cropland proportion, with some factor combinations yielding q-values &amp;gt; 0.9, confirming multi-factor control over carbon storage changes.[Conclusions] Carbon storage in the Guangxi sugarcane-producing region is shaped by land use patterns and multi-factor interactions. Future strategies should optimize land use structures, strengthen forest conservation, and balance urbanization with ecological protection to enhance regional carbon sequestration.

  PublisherOA PDFDOI

• Beyond the scenic view: a multimodal discourse analysis of sustainable tourism imaginaries on TikTok in Anhui, China

  Humanities and Social Sciences Communications · 2025-05-20 · 8 citations

  articleOpen access1st authorCorresponding

  Emerging tourism destinations harness digital technologies to reconstruct their tourism imaginaries and thereby gain global visibility. Nevertheless, there remains a limited understanding of how these destinations utilize multimodal resources to communicate their sustainable tourism imaginaries. Therefore, this study investigates how the Anhui Provincial Department of Culture and Tourism communicates its sustainable tourism imaginaries on TikTok. Drawing on the works of Wang and Feng and Feng, a semiotic framework is proposed to model Anhui’s imaginaries as evaluative attributes and analyze their construction through verbal and visual resources in short videos. The analysis of 203 TikTok videos reveals a positive representation of Anhui through six key imaginaries: natural beauty, vibrant cultural activities, contemporary chic, culinary experiences, traditional elegance, and rural idyll. These imaginaries reflect Anhui’s commitment to sustainability across social, environmental, and economic dimensions. By examining the multimodal communication of Anhui’s sustainable tourism imaginaries, this study provides insights for emerging destinations seeking to improve their visibility and appeal through sustainability-focused promotion.

  PublisherOA PDFDOI

• [Association Between Sleep Quality and Depressive Symptoms in Community-Dwelling Older Adults: The Mediating Role of Loneliness].

  PubMed · 2025-03-20

  articleOpen access1st authorCorresponding

  Objective: To examine the association between sleep and depressive symptoms among community-dwelling older adults and whether loneliness mediates this association. Methods: Using a multistage sampling approach, we enrolled participants aged 60 years or older from two communities in Chengdu, China. A questionnaire was used to collect basic information, including age, sex, etc., from the participants. In addition, loneliness, depressive symptoms, and sleep quality were assessed using a short-form University of California Los Angeles Loneliness Scale (ULS-8), the 10-item version of Center of Epidemiologic Studies Depression Scale (CESD-10), and the Pittsburgh Sleep Quality Index (PSQI), respectively. The Spearman rank correlation coefficient was employed to assess the correlations among social sleep, loneliness, and depression symptoms. Generalized structural equation modeling was used to assess the mediating effect of loneliness between sleep and depressive symptoms. Results: < 0.001). Conclusion: Poor sleep quality is associated with a higher risk of depressive symptoms in community-dwelling older adults, with loneliness mediating the association. Further research on improving the sleep quality to mitigate depressive symptoms in older adults is warranted. Special attention should be given to older adults experiencing both poor sleep and loneliness.

  PublisherOA PDFDOI

Frequent coauthors

• Barry M. Popkin

  91 shared

• Huijun Wang

  Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai)

  76 shared

• Bing Zhang

  Huzhou Central Hospital

  59 shared

• Penny Gordon‐Larsen

  50 shared

• Annette Peters

  Deutsches Diabetes-Zentrum e.V.

  47 shared

• Ching‐Yu Cheng

  46 shared

• Jost B. Jonas

  University of Washington

  46 shared

• Tien Yin Wong

  Singapore National Eye Center

  45 shared